A possible cause/answer to the big end failures ?
14 January 2002, 05:37 PM
#1
A few months ago, an engineer wrote an e-mail to the FAQ address enquiring about engine failures in Subaru Imprezas, and asking for more information. I supplied him with many links to topics on here and on 22b.com, plus as much information as I had at the time. I mentioned that I would be very interested in his findings (which was a full investigation into the cause of the failure). This may not be cause in all cases, but would certainly account for quite a few.
It would also explain why the Japanese get relatively few engine failures compared to us. As well as getting better quality fuel (which would help counteract det anyway), their cars are rarely run at high speeds for any length of time. Most still have the speed limiters in place, and those that have them removed would probably run their cars in the cold of night anyway.
UK cars (which are tuned for lower quality fuel and probably have more scope for adjustment), wouldn't suffer in the same way, as they also have a lower top speed.
Anyway, enough waffle - here are the findings:
"Apart from the results of the metallurgical tests, which indicate a compressive fracture, all the rest is done.
It would appear that the problem is one of aerodynamics.
When the vehicle is driven at high speed, the laminar airflow starts to lift above the bonnet line, which causes a partial depression at the front of the bonnet.
This in turn starves the intercooler of airflow, which causes its temperature to rise.
The result is an increased charge temperature that causes the onset of detonation.
The engine control module cannot make a sufficiently large adjustment to compensate for this, and so a percussion component sets up along the connecting rod, with the result that the Big End bearing shatters - more from the frequency than the force - and, as in this case, a compressive fracture of the rod results.
The solution is relatively simple - nothing much can be done about the aerodynamic shape without involving a seriously grown up cheque book.
However, a temperature sensor on the intercooler with appropriate warning light and/or buzzer in the cockpit will at least warn of the problem before any damage is done. "
[Edited by JohnS - 1/14/2002 5:39:59 PM]
It would also explain why the Japanese get relatively few engine failures compared to us. As well as getting better quality fuel (which would help counteract det anyway), their cars are rarely run at high speeds for any length of time. Most still have the speed limiters in place, and those that have them removed would probably run their cars in the cold of night anyway.
UK cars (which are tuned for lower quality fuel and probably have more scope for adjustment), wouldn't suffer in the same way, as they also have a lower top speed.
Anyway, enough waffle - here are the findings:
"Apart from the results of the metallurgical tests, which indicate a compressive fracture, all the rest is done.
It would appear that the problem is one of aerodynamics.
When the vehicle is driven at high speed, the laminar airflow starts to lift above the bonnet line, which causes a partial depression at the front of the bonnet.
This in turn starves the intercooler of airflow, which causes its temperature to rise.
The result is an increased charge temperature that causes the onset of detonation.
The engine control module cannot make a sufficiently large adjustment to compensate for this, and so a percussion component sets up along the connecting rod, with the result that the Big End bearing shatters - more from the frequency than the force - and, as in this case, a compressive fracture of the rod results.
The solution is relatively simple - nothing much can be done about the aerodynamic shape without involving a seriously grown up cheque book.
However, a temperature sensor on the intercooler with appropriate warning light and/or buzzer in the cockpit will at least warn of the problem before any damage is done. "
[Edited by JohnS - 1/14/2002 5:39:59 PM]
14 January 2002, 05:42 PM
#2
Scooby Regular
Join Date: Oct 1999
Posts: 3,491
Likes: 0
Received 0 Likes
on
0 Posts
Wow John, thanks for that bit of info !
Very interesting theory, and it would explain quite a few failures.
Does the engineer want to remain anonymous, or would he come out to play on Scoobynet ?
Theo [now wondering if anyone with a FMIC has ever suffered big end failure]
Very interesting theory, and it would explain quite a few failures.
Does the engineer want to remain anonymous, or would he come out to play on Scoobynet ?
Theo [now wondering if anyone with a FMIC has ever suffered big end failure]
14 January 2002, 05:50 PM
#3
Scooby Senior
Join Date: Oct 2001
Posts: 2,602
Likes: 0
Received 0 Likes
on
0 Posts
Interesting!
The question now is, has anyone with a front-mounted intercooler had this problem! That would put a nasty big stick in the works
I'm sure if this is right there is a simple aerodynamic solution - a small lip or rubber doobrie in the right place can do wonders. Actually, anyone with access to a fast private road (or wherever...) could presumably sellotape bits of string to the bonnet & watch them flap about, then when fast enough they would lift up rather than following the line of the bonnet...
well, it might work. Then stick blobs of blue-tack around the place & see if it makes any difference.
Can I appear on Blue-Peter now
The question now is, has anyone with a front-mounted intercooler had this problem! That would put a nasty big stick in the works
I'm sure if this is right there is a simple aerodynamic solution - a small lip or rubber doobrie in the right place can do wonders. Actually, anyone with access to a fast private road (or wherever...) could presumably sellotape bits of string to the bonnet & watch them flap about, then when fast enough they would lift up rather than following the line of the bonnet...
well, it might work. Then stick blobs of blue-tack around the place & see if it makes any difference.Can I appear on Blue-Peter now
14 January 2002, 05:57 PM
#4
Scooby Regular
Join Date: Dec 2000
Posts: 2,912
Likes: 0
Received 0 Likes
on
0 Posts
A really interesting theory, and not too far fetched either...I can see me savings (what's left of 'em ) shot to hell in a not too distant future...
I think I needed that FMIC anyway...
/J
) shot to hell in a not too distant future...I think I needed that FMIC anyway...
/J
14 January 2002, 06:52 PM
#5
Scooby Regular
Join Date: Jul 2000
Location: Glasgow
Posts: 3,040
Likes: 0
Received 0 Likes
on
0 Posts
That's interesting reading John.
I was told recently by persons unnamed, but associated with a certain Subaru tuning arm, that benefits to intercooler flow could be made by moving the number plate out so that it was flush with the bumper. The reasoning being that the air flow in this area was disturbed by the plates position and reduced the effectiveness of the intercooler because the air was turbulating around the lower half of the bumper and hence not flowing smoothly into the scoop.
This mod may be effective for a competition car where tenths of a second are important in a stage or a lap but felt it wouldn't be much use for everyday driving.
At first i thought it was a **** take, but now i'm not so sure.
A
P.S. This was recommended to me cause i had the '99 year car. Don't know about others.
I was told recently by persons unnamed, but associated with a certain Subaru tuning arm, that benefits to intercooler flow could be made by moving the number plate out so that it was flush with the bumper. The reasoning being that the air flow in this area was disturbed by the plates position and reduced the effectiveness of the intercooler because the air was turbulating around the lower half of the bumper and hence not flowing smoothly into the scoop.
This mod may be effective for a competition car where tenths of a second are important in a stage or a lap but felt it wouldn't be much use for everyday driving.
At first i thought it was a **** take, but now i'm not so sure.
A
P.S. This was recommended to me cause i had the '99 year car. Don't know about others.
14 January 2002, 07:01 PM
#6
Scooby Regular
Join Date: Nov 2001
Posts: 1,072
Likes: 0
Received 0 Likes
on
0 Posts
Interesting findings.
What exactly would be classed as "high speed" ?
If Japan is not experiencing the problem I guess >112mph ?
Presumably this is also long blasts at high speed. i.e a quick thrash up to say 140 (on private land of course) then returning to <100 would I guess not raise the temp too much.
All of this is guessing of course, but any extra info might be useful for those who want to keep their engines intact.
What exactly would be classed as "high speed" ?
If Japan is not experiencing the problem I guess >112mph ?
Presumably this is also long blasts at high speed. i.e a quick thrash up to say 140 (on private land of course) then returning to <100 would I guess not raise the temp too much.
All of this is guessing of course, but any extra info might be useful for those who want to keep their engines intact.
14 January 2002, 07:10 PM
#7
My interpretation is that it would be high speed running for an extended period (perhaps 15 minutes, perhaps longer ?), so the intercooler temps rise over a period of time. The car would presumably have to be going at a certain speed, where the amount of work it needs to do is significantly greater as well, and therefore running hotter.
I'm not sure of these things, but perhaps the effects could be cumulative over time (like metal fatigue), and it could take repeated extensive runs at high speed to cause problems.
I'll contact the engineer to see if he wishes to add anything to the discussion. I did ask if I could forward on the information from his findings, and was given permission to do so, in the hope that it may save a few people from an expensive rebuild.
I'm sure someone posted on here a while back, that the EGTs rose suddenly above a certain speed. Perhaps this is the critical speed where airflow into the intercooler is disrupted enough to have a significant effect, causing the charge temps to rise ? Of course, there are probably other factors involved as well, such as ambient temperature and wind speed/direction ?
I'm not sure of these things, but perhaps the effects could be cumulative over time (like metal fatigue), and it could take repeated extensive runs at high speed to cause problems.
I'll contact the engineer to see if he wishes to add anything to the discussion. I did ask if I could forward on the information from his findings, and was given permission to do so, in the hope that it may save a few people from an expensive rebuild.
I'm sure someone posted on here a while back, that the EGTs rose suddenly above a certain speed. Perhaps this is the critical speed where airflow into the intercooler is disrupted enough to have a significant effect, causing the charge temps to rise ? Of course, there are probably other factors involved as well, such as ambient temperature and wind speed/direction ?
Trending Topics
14 January 2002, 07:19 PM
#8
Scooby Regular
Join Date: Nov 2001
Posts: 1,072
Likes: 0
Received 0 Likes
on
0 Posts
This could explain things (for me at least).
Having been tuned to 348BHP (from 284 originally) and living in dread of some failure ever since, spending a fortune on Oil Coolers and Water injection for safety, but now having made it to 57K miles (touch wood).
Although I drive my car hard, I always let it warm up and cool down appropriately, however I rarely go >100 because of licence loss implications, and when I do it's just for 1/2mile or so.
On top of which I've never taken my WRX out of the UK, so I've not been tempted to thrash it max out for any period.
I remember Anders (just an example, nothing personal mate) posting to a web site or newsgroup about how he'd maxed out his 22B in Germany (or maybe France) for a long period of time, and I heard he has had some engine failure problems in the past.
This indeed is interesting info.
Maybe Scoobynet could conduct a pole of those who regularly exceed big MPH and how many have had engine failure (making sure of course to keep the information safe from the boyz in blue).
Having been tuned to 348BHP (from 284 originally) and living in dread of some failure ever since, spending a fortune on Oil Coolers and Water injection for safety, but now having made it to 57K miles (touch wood).
Although I drive my car hard, I always let it warm up and cool down appropriately, however I rarely go >100 because of licence loss implications, and when I do it's just for 1/2mile or so.
On top of which I've never taken my WRX out of the UK, so I've not been tempted to thrash it max out for any period.
I remember Anders (just an example, nothing personal mate) posting to a web site or newsgroup about how he'd maxed out his 22B in Germany (or maybe France) for a long period of time, and I heard he has had some engine failure problems in the past.
This indeed is interesting info.
Maybe Scoobynet could conduct a pole of those who regularly exceed big MPH and how many have had engine failure (making sure of course to keep the information safe from the boyz in blue).
14 January 2002, 07:47 PM
#9
Scooby Regular
Join Date: Aug 1999
Posts: 1,561
Likes: 0
Received 0 Likes
on
0 Posts
high charge temps have always known to cause det. i remember when i tested the charge temps the do tend to rise pretty, pretty quickly. in 5th gear lets say above 5000rpm it will only take couple of minutes to get you in the 60 degrees if not more.
very interesting thread. though
sam
very interesting thread. though
sam
14 January 2002, 08:57 PM
#10
I have noticed in the past that at high speed water from the intercooler spray finds it's way onto the screen and bonnet. Maybe it needs a larger (taller) scoop like the new UK STi. Maybe that's why they use it - probable P1 lessons learned?
[Edited by Iain P - 1/14/2002 8:58:23 PM]
[Edited by Iain P - 1/14/2002 8:58:23 PM]
14 January 2002, 09:09 PM
#11
Scooby Senior
Join Date: Oct 2001
Posts: 2,602
Likes: 0
Received 0 Likes
on
0 Posts
I still think we should tie bits of string to the bumper to see if it really happens
On a more important note (& presuming that this thoery is correct) it would be very useful to know at what speed the separation occurs. For anyone who either lives in or is passing through the areas of Europe where 130+ish is the comfortable speed to be travelling at for long periods, what is the safe maximum speed? This is for those of us who don't want to start slapping sensors all over the place that may or may not give correct readings (as it's unlikely they were designed for this sort of thing...) I still recon string would be the best indicator - cheap, adjusts to wind conditions, temperature, etc.
On a more important note (& presuming that this thoery is correct) it would be very useful to know at what speed the separation occurs. For anyone who either lives in or is passing through the areas of Europe where 130+ish is the comfortable speed to be travelling at for long periods, what is the safe maximum speed? This is for those of us who don't want to start slapping sensors all over the place that may or may not give correct readings (as it's unlikely they were designed for this sort of thing...) I still recon string would be the best indicator - cheap, adjusts to wind conditions, temperature, etc.
14 January 2002, 09:23 PM
#13
Scooby Regular
Pete Croney and I were talking about this recently, and is *exactly* the conclusion he came to. And no, it's not Pete trying to flog me an STI scoop, because i've already got one on order
The base of the windscreen is a known dead/low pressure area - just look how the washer water/leaves just hang around there - and many of the big end failures appear to have been associated with high speed running - when the airflow into the I/C scoop is probably stalled the most.
[Edited by Dave T-S - 1/15/2002 9:49:03 AM]
The base of the windscreen is a known dead/low pressure area - just look how the washer water/leaves just hang around there - and many of the big end failures appear to have been associated with high speed running - when the airflow into the I/C scoop is probably stalled the most.
[Edited by Dave T-S - 1/15/2002 9:49:03 AM]
17 January 2002, 01:25 PM
#14
Scooby Regular
Join Date: Nov 2001
Location: The Emerald Isle
Posts: 869
Likes: 0
Received 0 Likes
on
0 Posts
hi just to add me 2ps worth, as a scoobie newbie this big end failure thread made me s**t my pants, im sure u've all read the large thread on the no3 big end failure (bob rawle etc) IMHO alot of the failures were on cars which had just completed sustained hi speed runs, and the failure seemed to occur when the cars slowed down, as a previous Toyota owner (not a gt4 a humble celica 4ws ;-)) i read alot of gt4 posts, the majority of AtoA intercoolers were fitted with pace electric fans to reduce intake temp, is this something which may benefit the scoob?
Cheers Fergal
Cheers Fergal
17 January 2002, 01:54 PM
#16
Scooby Senior
Join Date: Oct 2001
Posts: 2,602
Likes: 0
Received 0 Likes
on
0 Posts
I think the problem is that no air is getting to the scoop because of the laminating flow, and the water spray cooling down 'no air' would probably just make things wet. The air still isn't there!
On the fan thing, one gets a mention here as making a little bit of difference, but I think it was only tested at normal driving speeds rather than at the speeds at which the airflow starts to go the wrong way. It may be the case that the little bit of suction it creates would be enough to disturb the flow of air so that it doesn't delaminate, but it seems unlikely unless those fans are pretty serious!
On the fan thing, one gets a mention here as making a little bit of difference, but I think it was only tested at normal driving speeds rather than at the speeds at which the airflow starts to go the wrong way. It may be the case that the little bit of suction it creates would be enough to disturb the flow of air so that it doesn't delaminate, but it seems unlikely unless those fans are pretty serious!
17 January 2002, 01:59 PM
#17
Scooby Regular
Join Date: Jul 2001
Posts: 215
Likes: 0
Received 0 Likes
on
0 Posts
If the failures are a result of det due to high intake temps, I assume this means the ECU has retarded the ignition as far as poss but det still continues. Assuming the factory knock sensor is functioning, would the ECU not go into "limp home" mode - ie restrict revs to 2500. or it the ECU not that intelligent?
Dave
Dave
17 January 2002, 02:00 PM
#18
Scooby Regular
Join Date: Nov 2001
Location: The Emerald Isle
Posts: 869
Likes: 0
Received 0 Likes
on
0 Posts
interesting link on the fans, AFAIR the gt4 lads mounted theres on top of the intercooler... but as i think about this the airflow over the celica would appear to be more efficent than the scoobie, also the later gt4's "scoop" was more of a vent (open grill) which maybe helped too??? has anyone vented the scoobs bonnet near the intercooler? would utillising the redundant vents and routing air to the intercooler help?
Cheers Fergal
Cheers Fergal
17 January 2002, 02:13 PM
#19
Scooby Regular
Join Date: Oct 1998
Location: Scoobysport, Basildon, UK
Posts: 4,300
Likes: 0
Received 0 Likes
on
0 Posts
John I have been dealing with Douglas (the engineer in question) for some time with this. He is investigating an engine failure warranty claim. He has not done any research on this and the contents of his mail are my theory, not proven fact.
Although the mail he sent you was just a summary of my thoughts, I have been discussing this theory with Bob Rawle and others for quite a long time.
There is another thread about intercooling and I have put more of my thoughts there. http://www.scoobynet.co.uk/bbs/threa...D=65130&Page=1
Lets keep this one going with discussion about this though.
Although the mail he sent you was just a summary of my thoughts, I have been discussing this theory with Bob Rawle and others for quite a long time.
There is another thread about intercooling and I have put more of my thoughts there. http://www.scoobynet.co.uk/bbs/threa...D=65130&Page=1
Lets keep this one going with discussion about this though.
17 January 2002, 02:16 PM
#20
Front mounts tend to be a favoured modification here in Oz.
There are numerous manufacturers of this component.
I stuffed about with splitters and sprays -Never convinced they really contributed to lowering intake temperatures that much.
Now with the Hyperflow FMIC in place I am confident that even on the hottest days a safe intake temp is being maintained .
Has to be said that its crap subaru design though.
There are numerous manufacturers of this component.
I stuffed about with splitters and sprays -Never convinced they really contributed to lowering intake temperatures that much.
Now with the Hyperflow FMIC in place I am confident that even on the hottest days a safe intake temp is being maintained .
Has to be said that its crap subaru design though.
17 January 2002, 02:30 PM
#21
Scooby Regular
Join Date: Oct 1999
Posts: 3,491
Likes: 0
Received 0 Likes
on
0 Posts
Bob, I hope you don't mind quoting this from a 22B com thread that went a bit off topic later on
I've actually measured the airflow thro' the intake over a wide range of speeds, there is no reduction at any time only the expected increase in flow as velocity increases. Pete C told me about the theory which is why I did the check, I tested my car using a flow meter mounted in the centre of the scoop. I monitor charge temps as well, the rate of increase in temp with any top mount did not appear to change with velocity except to slow (again as expected). I used a Motec ecu to datalog this.
Its a sound theory though but was it proven? Maybe its MY dependant or maybe my car is just not quite the right shape to cause the problem. I've been convinced about the speed limit theory though for some time, the jap spec cars were definately designed "UP" for the japanese market which is speed limited meaning never getting to high revs in 4th or 5th.
The plot thickens, as they say.
Are there any windtunnels available where we could test a few cars / MY's without breaking the bank ? Or could we use said flow meter on a few cars ?
I've actually measured the airflow thro' the intake over a wide range of speeds, there is no reduction at any time only the expected increase in flow as velocity increases. Pete C told me about the theory which is why I did the check, I tested my car using a flow meter mounted in the centre of the scoop. I monitor charge temps as well, the rate of increase in temp with any top mount did not appear to change with velocity except to slow (again as expected). I used a Motec ecu to datalog this.
Its a sound theory though but was it proven? Maybe its MY dependant or maybe my car is just not quite the right shape to cause the problem. I've been convinced about the speed limit theory though for some time, the jap spec cars were definately designed "UP" for the japanese market which is speed limited meaning never getting to high revs in 4th or 5th.
Are there any windtunnels available where we could test a few cars / MY's without breaking the bank ? Or could we use said flow meter on a few cars ?
17 January 2002, 02:38 PM
#22
Scooby Regular
Join Date: Oct 1999
Posts: 3,491
Likes: 0
Received 0 Likes
on
0 Posts
Pete, just so I interpret things correctly: in the other thread you say:
1) We have done engine rebuilds/swaps on over 80 Imprezas. Most are STi, many are WRX, 2 were UK cars. None had front mounts.
In your opinion, would that be because UK cars fail less often, or because UK owners would be inclined to go to their Subaru dealer (warranty etc )? I have heard of 4 cases with UK cars (that I know of) in the Netherlands, some were tuned, some were bog standard.
On one car with big end failure, there were no visible traces from det at all. Could it be that some types of det don't leave those marks ?
Theo
1) We have done engine rebuilds/swaps on over 80 Imprezas. Most are STi, many are WRX, 2 were UK cars. None had front mounts.
In your opinion, would that be because UK cars fail less often, or because UK owners would be inclined to go to their Subaru dealer (warranty etc )? I have heard of 4 cases with UK cars (that I know of) in the Netherlands, some were tuned, some were bog standard.
On one car with big end failure, there were no visible traces from det at all. Could it be that some types of det don't leave those marks ?
Theo
17 January 2002, 07:34 PM
#23
Scooby Regular
Join Date: Jun 1999
Posts: 1,430
Likes: 0
Received 0 Likes
on
0 Posts
Autospeed did some "wool tuft" testing of the MY99 WRX which may be of interest:
http://www.autospeed.com/A_1055/page1.html
Thanks
Gavin
http://www.autospeed.com/A_1055/page1.html
Thanks
Gavin
17 January 2002, 08:22 PM
#24
Scooby Regular
Join Date: Oct 1998
Location: Scoobysport, Basildon, UK
Posts: 4,300
Likes: 0
Received 0 Likes
on
0 Posts
Theo
I suspect its a number of reasons.
1) We will only see cars that are out of warranty.
2) The European cars are running at low boost pressure at very high speeds, so less heat is generated.
The STi/WRX turbos will still be pulling a bar of boost at 140mph, the EU car would have about 0.5 bar.
I suspect its a number of reasons.
1) We will only see cars that are out of warranty.
2) The European cars are running at low boost pressure at very high speeds, so less heat is generated.
The STi/WRX turbos will still be pulling a bar of boost at 140mph, the EU car would have about 0.5 bar.
17 January 2002, 10:22 PM
#25
Ecu Specialist
This has awakened my interest again so I will do some more checks, alot of the feedback I get about big end failure points to this happening after a "blast" and then a slow down with the problem becoming apparent when the car is asked to go again, this could be oil related and, in virtually every case, the cars were serviced withinn 800 miles of the problem, most within 100 miles.
The delamination theory is logical, if charge temps are being monitored then this would be apparent, its possible that cars are more or less prone to this due to the way they attack the air flow, I have always monitored charge temps and always backed off once I got over 50/55 degrees with the top mount, became a habit.
I'll see what results I get when I get the chance to try this again.
The delamination theory is logical, if charge temps are being monitored then this would be apparent, its possible that cars are more or less prone to this due to the way they attack the air flow, I have always monitored charge temps and always backed off once I got over 50/55 degrees with the top mount, became a habit.
I'll see what results I get when I get the chance to try this again.
18 January 2002, 06:52 AM
#27
Scooby Regular
Join Date: Jan 2002
Posts: 81
Likes: 0
Received 0 Likes
on
0 Posts
Most were STi. Hrm. Perhaps the STi ECU can't pull as much timing or the average number of revs is higher due to the 4.44 gearing. I assume all STi models are geared this way. Now I'm starting to get scared about this JDM swap I'm doing.
-Michael
-Michael
18 January 2002, 09:57 AM
#28
Scooby Regular
Join Date: Oct 1999
Posts: 3,491
Likes: 0
Received 0 Likes
on
0 Posts
Just for your info ... I have asked the German Impreza club if they know of some/many cases involving the N°3 big end failure. I reread the whole original thread this morning, and thought that if it is related to high speed blasts, the Germans might see more failures due to having long stretches of autobahn with no speed limits.
I'll keep you posted on the progress of that thread, and I also linked back to this thread.
Theo
I'll keep you posted on the progress of that thread, and I also linked back to this thread.
Theo
18 January 2002, 11:50 AM
#29
Scooby Regular
Join Date: Nov 2000
Posts: 3,675
Likes: 0
Received 0 Likes
on
0 Posts
When mine was stripped down, there didn't appear to be any real sign of det......
One thing I did notice though - how much the bonnet "arcs" at high speed - looking like it's got tremendous pressure pushing up....
This is all definitely a case for gauges....if I'd had inlet temp, oil temp or pressure and perhaps egt, I would've backed off when I saw what was happening.
One thing I did notice though - how much the bonnet "arcs" at high speed - looking like it's got tremendous pressure pushing up....
This is all definitely a case for gauges....if I'd had inlet temp, oil temp or pressure and perhaps egt, I would've backed off when I saw what was happening.